Is it worth developing a sub £100 3D printer not evolved from current designs?

I have previously posted on another hobby development related to Reprap that I have worked on and I felt there was a general feeling that people should not try and leap forward on different forks or tangential concepts and should concentrate only on evolutionary minor improvements of the existing Reprap systems.

I have been thinking about a fairly radical design of machine for a few months, thought I have not put anything on paper or made any solid calculations I am guessing that a sub £100 3D printer could easily be made with existing off the shelf parts (accuracy and speed may be an issue though).

I have really lost interest in playing with this stuff and have moved onto other interests recently, but am still thinking maybe it is worth doing next time I am at a loss for an interesting project.

I wonder if the community have any opinion about whether posting details on the reprap forum of systems that may paradigm shift the existing methods by which printers are constructed would adversely effect the community?

I don't want to start posting informations and have people feel that I am confusing and splitting development or have people misunderstanding the design considerations and feeling disinformation has been posted.

I highly respect the Reprap community and do not want to snub anyone.

As said I am not working on this at present but may in a year or so depending how things go, so really I hoped to get some feedback in case I decide to post on the Reprap forum if and when I start development.

Building a sub £100 would be really low. On the wiki a cost reduction page estimates a $300 (about £190)
( [reprap.org] ). It all depends on your box of available/salvaged parts and toolbox.

You will probably be building a repstrap ( [reprap.org] ). Problem with development of a cheap printer based on your own available parts is reproducability. Your collection is most likely unique and will be much more expensive and harder to source for others wanting to build a similar printer based on your hard development hours.

If you want others to gain from your radical design you may want to think about printable alternative sections and easy to source parts. Perhaps some ikea chair costing $15 can be used to create 80% of the frame with just minor modifications. This would be reproducable and easy to source (for the next 5 to 15 years).

I have been looking at cheap designs too. If Canon can sell a laser printer for less than $50, it should be possible to build a reprap under $150 without scavenging for thrown-away parts.

A few metres of 8mm rod - $20
Some bearings at ~$1 ea (e.g. 10)
Cheap Arduino $20
Printed frame
Printed bushings

That leaves a budget for motors, hot-end and drive electronics of ~$80, which is certainly possible.

I tried using RC servos to replace steppers, but they are a bit too slow and inaccurate even for closed-loop positioning.

Some other things I'm planning to try are:

Unipolar steppers driven by transistor arrays
Printed steppers (NdFeB magnets, ferromagnetic powder and copper wire)
Cheap gear-head steppers.
Brushless DC motors with some form of positioning feedback.

It would be nice if a hot-end could be printed, or made from a printed mould. Maybe a design with slots instead of threaded connections could work.

If you have some ideas, you might as well post them and get some feedback which you can take into consideration.

mung: I do not think you could possibly post any information that would make the community any more fractured or chaotic than it already is. I think your idea sounds interesting, and many other people here probably feel the same way. I say post it!

darkpaw: It makes me really happy to see other people working on alternatives to steppers. I would really like to see a fully printable RepRap someday - just because I think it would be cool. (Hence my excitement about printable motors). Just in case you haven't seen this stuff yet, there are some printable and quasi-printable motors on thingiverse, and there is a review page on possibly-printable actuators on the wiki:
solenoid motor
printed coil motor
Actuator Fabrication Review

If reduced cost is the main consideration, one might be able to use cheap DC motors, with printed gearboxes, printed encoder wheels, and cheap drive electronics on a custom reprappable pc board. This combination would probably cost less than current alternatives.

But I have to wonder: There are so many players now in the low-cost 3D printer business, it seems like if a sub £100 printer were possible, someone would be selling it ( or at least pushing it on Kickstarter )

Quote

I wonder if the community have any opinion about whether posting details on the reprap forum of systems that may paradigm shift the existing methods by which printers are constructed would adversely effect the community?

"Adversely"? Like unfriendly competitors will speak up? Of course they will, but that's more a proof you found something interesting than a problem. Just keep also in mind people do something the way they do it for a reason. So only few will change their way to build a printer just for the sake of doing something differently.

Regarding documenting your ideas, I think the best way is to put new ideas into the wiki. There you can edit the stuff, add pictures (very important) and people will find it half a year later, too. For discussing these findings, this forum is about the best we have. Here one can take some time to research an answer, to think a bit, to write well founded answers. There's also IRC, but on IRC, well, you meet only a few people which happen to be there at that time and a few minutes after starting, a topic is already lost.

Regarding general designs ... AFAIK, a WolfStrap is still the cheapest design currently available. It uses a wooden frame and generally wasn't even accepted to be a "true" RepRap for this. Now, two years later, a brand new "Prusa Mendel i3" design comes around the corner, is cheered as the best and truest RepRap ever and it uses -- o-oh -- a wooden frame.

Now, will you enlighten us where we waste money, please? Honestly, I don't see a way to make the electronics as well as mechanics cheaper without substantial loss of usability. Oh wait ... on Amazon I've seen drawer slides for EUR 1.50 a pair, which make ... now, what are your ideas?

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

As said I am not working on this at present but may in a year or so depending how things go, so really I hoped to get some feedback in case I decide to post on the Reprap forum if and when I start development.

A year is a long time, I would ask the question when you are ready to go, but if you decide to develop it then go for it, and although the forum is getting more confrontational by some members I hope the majority are still supportive of new development.

The fact that you feel the need to ask is worrying, this should be a place for ideas and development and should not be dictated by anyone.

---

Random Precision

I would post any ideas you have about the design as there are many here who would be interested. A forum is for sharing ideas and getting inspiration from others. Just ignore the "It won't work" comments. History is littered with genius ideas that have been delayed by critics.

Its not useful to compare the price of a printer from Canon etc. to that of a reprap, as many printers are sold as Loss Leaders. This is when they are sold at cost or at a loss and the costs recovered by the sale of the printer cartridges. This is why each printer maker uses their own brand of cartridge and not a standard one. A further confusion is that China controls its exchange rates. This is why you can buy a DVD player for £15 in Tescos even though the parts alone actually cost more.

Having said all that £100 does sound a bit cheap as the main cost is in motors and electronics. We would be keen to hear your strategy for this part of the design. Even if you import most of your hardware, import duty to the UK could eat a large chunk of your £100.

Edited 2 time(s). Last edit at 06/19/2012 08:44AM by martinprice2004.

The economics of scale indicate that stepper driver + motor are the first point to look at in a reprap.

A Stepper 44ncm Stepper can be sourced for 20$
The everpopular stepstick goes for ~ 7$
So you spent 27$ per axis.
You got 4 of Them (XYZ+E).
So thats 125$(in a 5 motor prsua) if you are sourcing Jedi.
For most People 150$+ are much more realistic.


The current Mendel Version is a nice tradeoff between system complexity,cost and performance.

If you want to archive the same performance at lower cost there is a high possibility you might have to increase system complexity. You can do so by using Brushless motors with encoders.
Sidenote: Brushless Motors are the simplest motors to build, and do not require iron in slotless configuration.So if you plan on building your own motors thats the first direction to look.By Building your own Motors and discrete servocontrollers you can push the per-axis price well below 10$.
But even if you just go out and buy our motor in china you may be able to scratch the 11-12$ mark per axis.

3$ Encoder
1$ uC
2.50$ for 6xN-fet + predriver
5$ for a rc brushless motor

So you end up with 60$ complete smd electronics.

I guess you could do it even cheaper if you searched for more than 20 minutes.
By building your own montorcontroll solution and a motor system complexity will skyrocket compared to a simple step+direction pulolu solution. But it allowed you to cut the electronics price in half.

This leaves you with a 40$ Budget for Hardware which should be enought for a 10 lm8uus + a few meters of extrusions and smoothrods (i did not do the math on this one)

(to be continued...)

The cost to manufacture a copy of complex software is exactly the same as for simple software. Sure it might take a bit more development effort, but it only has to be done once.

On the hardware side though, losing the complex and proprietary stepper driver chips can only be a good thing IMO.

But on the other hand, maybe it's worth looking at printing millimetre scale semiconductor devices to solve that problem. Peltier devices have decent quantities of N and P type semiconductor material, I wonder how badly you can treat it before it fails, or how good the junctions have to be to make e.g. a transistor?

When I said the price of sub £100 I meant using readily available off the shelf components.

I was thinking of using these :

[www.ebay.co.uk]
or
[www.ebay.co.uk]

Crap from china, but the magic is in getting a usable system from such low powered and torqueless motors.

Basically almost everything will be made from very light weight plastic that could be self replicated.

I did also toy with the idea of using a homemade system of solenoids to drive the machine as they are very easy to make and would remove much of the need for drive electronics.

I cannot tell the system yet as I have not investigated it yet, but it requires some very complex hyperbolic maths to calculate the kinematics, and the controller would be a raspberry pi type device.

The major problem could be in the cost of plastic though, as I am guessing around 1.8kg would be needed?

I doubt I will ever get time to do it though, but I think it is possible. Maybe when I get time I could elaborate further.

Why not look at repurposing something that already has most of the same goals. A cheap inkjet printer. Pick one that is new on market ( so its around for a while) and under $60.

It already has two steppers, xl belt, pulleys, shaft, power supply, carraige assembly, a control board, limit switches.

Or an older one of the popular all in ones would add two more steppers and second shaft (scanner and top feeder).

Bob